Brain storming
obecity
overweight
unhealthyness
healythyness
intelligence
marks
grade point average
schools


healthyness effect intelligence?
yes
what affects marks
idoes inteligence effect marks
yes and very subjective
does healthyness effect marks
healthyness too hard to define
obecity effect individual marks?
individual marks too hard to find data on
individual obecity rates too hard to find data on


Question

Are the average marks of schools affected by the average students in the school that are overweight.

begginning Hypothesis

I believe that schools with a high average amount of students that are overweight will, have a lower grade point average.

PHASE 2

links thus far

http://www.nature.com/oby/journal/v14/n10/fig_tab/oby2006212t2.html#figure-title

http://www.nature.com/oby/journal/v14/n10/fig_tab/oby2006212t1.html#figure-title

http://www.econstor.eu/bitstream/10419/27310/1/571528201.PDF

Hypothesis: I believe that there will be a moderate positive correlation between obescity and grade point average.

Reasoning: I believe this will be because people that are obese have a higher chance of heart disease along with a variety of other serious medical conditions. These medical conditions could have a serious effect on concentration and ability to preform academically.

Potential Bias: There is a strong possibility for sampling bias, along with extrenuos variables. Sampling bias could come from only taking samples from one culture, or area of people, extrenuos variables could come from things such as poverty and general culture of the people involved.

New topic

Question: does being overweight / obese effect the perceived mental health of a person?

hypothesis: I think that being overweight / obese will strongly negatively effect the mental health of a person.



2003
2005
2007
2008
2009
NF





Good Mental
72.2
79.8
75.2
77.5
70.7
Poor Mental
n/a
n/a
n/a
n/a
n/a
Overweight / obese
26.3
36.7
46.7
35.0
47.3
PEI





Good mental
76.0
76.8
67.1
78.3
73.2
Poor mental
n/a
n/a
n/a
n/a
n/a
Overweight /obese
28.8
43.0



NS





Good mental
79.0
80.2
76.8
74.3
75.3
Poor mental
7.3

5.4


Overweight /obese
28.3
33.3
n/a
26.3
29.0
New Brunswick





Good mental
76.2
76.5
77.7
75.9
71.8
Poor mental
4.5
2.5
n/a
7.7
n/a
Overweight /obese
24.3
32.2
34.3
27.7
25.6
Quebec





Good mental
80.2
77.4
80.7
83.1
78.1
Poor mental
2.5
2.7
2.5
1.7
1.7
Overweight /obese
16.8
21.3
18.3
27.2
22.5
Ontario





Good mental
n/a
76.2
76.3
77.4
77.4
Poor mental
n/a
3.4
4.5
4.3
3.7
Overweight /obese
n/a
22.5
20.9
27.2
25.4
Manitoba





Good mental
76.8
76.6
76.6
74.9
81.6
Poor mental
2.8
3.4
n/a
n/a
4.2
Overweight /obese
22.8
17.3
39.9
30.1
29.9
Saskatchewan





Good mental
75.2
76.0
78.3
78.1
76.8
Poor mental
4.0
3.5
2.4
2.2
3.7
Overweight /obese
22.4
31.8
27.4
28.7
20.7

Alberta





Good mental
77.1
77.4
78.9
75.9
81.5
Poor mental
4.3
4.9
5.5
2.8
3.3
Overweight /obese
21.7
23.7
24.0
17.2
25.4
British Columbia





Good mental
71.9
77.5
75.1
80.9
73.8
Poor mental
4.1
4.8
3.5
2.2
3.1
Overweight /obese
18.8
24.2
17.2
20.0
27.2

The mental health in good condition data is has surprisingly little deviation at 2.90

the obesity / overweight data however had a higher standard deviation at 7.12


The correlation coefficient of this data is -0.23522, this is obviously much lower than I expected.
This is probably due to this being a study of the full population rather than a controlled group, this allowed bias
and large numbers of more "average" results to skew the data.



Picture3.jpg

Picture4.jpg



Bias:
There is a great deal of opportunity for this data to be biased. As I said before this data is from a census,
rather than a centralized study, this allows for a strong non response and response bias.
This is because in the census there is no motivation to answer truthfully or at all.
Beyond that bias there is also the fact that there are a huge number of variable that go into a persons mental health.
For this data to be accurate a full study would have to be performed with not a greater audience (being that this is a census)
but a smaller one, for greater control and to limit the extraneous variables that are present currently.

Conclusion:
My hypothesis was incorrect and with a correlation coefficient of -0.23522 my data is ultimately inconclusive.
This is because the coefficient is relatively low and there is a relatively high possibility of bias.


Source: Statistics Canada. Table 105-0501 - Health indicator profile, annual estimates, by age group and sex, Canada, provinces, territories, health regions (2007 boundaries) and peer groups, occasional, CANSIM (database).
http://cansim2.statcan.gc.ca/cgi-win/cnsmcgi.exe?Lang=E&CNSM-Fi=CII/CII_1-eng.htm
(accessed: January 16, 2011)